Ultimate PocketIA
IA
Actualité : Kimi K3 : la plus grosse IA open source et gratuite jamais publiée est disponible
© Moonshot AI - Le visuel officiel de Kimi K3, le modèle d'IA open source de Moonshot AI aux 2 800 milliards de paramètres. Yang Zhilin, cofondateur et PDG de Moonshot AI, n'a pas choisi sa date au hasard. Dévoilé le 16 juillet lors de la World Artificial Intelligence Conference de Shanghai aux côtés du Qwen 3.8 d'Alibaba, dans une double offensive chinoise sur l'IA open source, Kimi K3 est accessible via API depuis…
Lire l’article complet
IA
Apple's smart home hardware refresh draws closer with Siri AI smarts
Apple is going to make a bigger play for the smart home using Siri AI, with updates to the HomePod mini, a new Apple TV, and the rumored Home Hub arriving between October and early 2027. Apple has long been rumored to be working on a swathe of new smart home hardware, with Siri AI at the heart of the company's renewed home push. As Siri AI becomes a bigger figure in its main…
Lire l’article complet
IA
iOS 27’s Siri gets one thing very right that other AI assistants get wrong
Siri AI is a major upgrade to Apple’s assistant that’s coming in iOS 27. And there’s one standout area I’ve noticed where it differs greatly from other AI assistants—for the better. I’ve been using Siri AI for nearly two months in the iOS 27 beta. And so far, it’s checked all the boxes I was hoping for. The new Siri is far more capable than its predecessor, delivering on the original promise of being…
Lire l’article complet
IA
Fish Audio raises $52M seed to build AI voice models for creators and enterprises
The market for AI-generated voice models is massive. Creative use cases require AI voice models to be more expressive, while enterprises looking to automate customer support and sales ops need them to be more steerable. Palo Alto-based Fish Audio wants to cater to all of those use cases with its library of more than 15,000 natural language controls. Since launching last year, the startup now has more than 8 million people using the open…
Lire l’article complet
IA
Granola meeting notepad comes to watchOS as Apple Watch becomes an AI device
Granola, the AI-powered meeting notepad best known for its Mac app, is expanding to Apple Watch today. The watchOS app is designed for conversations away from a computer. Think sessions like walking one-on-ones, whiteboard sessions, and coffee meetings. One tap starts note-taking, with a green screen and sound confirming that Granola is listening. Scheduled meetings can trigger a subtle reminder, while a watch face complication puts recording one tap away. On the Mac, Granola…
Lire l’article complet
IA
5 ways AI Mode in Search helps you enjoy the real world
Here’s how Search's AI tools can help you make the most of your time offline, whether you want to book concert tickets or find the perfect hiking boots. Your browser does not support the audio element. People are searching for more and more ways to get offline, experience the real world, and “touch grass.” This year, people are searching for “adult tennis lessons,” “trail running,” “hiking,” and “run club” more than ever before. People…
Lire l’article complet
IA
5 ways to host the ultimate dinner party with Google Search
These AI features can help you craft a menu, design a tablescape, and handle other party-planning tasks. Your browser does not support the audio element. Planning the perfect dinner party is an art. Even if you love to host, it can be stressful to navigate logistics like crafting menus and designing tablescapes. Luckily, Search can help. As more people are looking to find ways to connect in real life, dinner party-related queries are also…
Lire l’article complet
IA
The Download: OpenAI’s predictable hack, and an AI stock sell-off
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. OpenAI called the Hugging Face attack unprecedented. But we’ve been here before. —Will Douglas Heaven, senior AI editor Reading OpenAI’s account last week of how some of its models broke their containment and hacked into the computer systems of Hugging Face, another AI company, was the first time I got…
Lire l’article complet
IA
Data centers IA : Nvidia envisage un soutien financier record de 250 milliards de dollars à OpenAI
Nvidia et OpenAI discutent d’un soutien financier pouvant atteindre 250 milliards de dollars pour un campus de data centers dans l’Ohio. Ce mécanisme permettrait à OpenAI d’emprunter en s’appuyant sur la solidité financière de Nvidia. La course à l’infrastructure IA prend une nouvelle dimension : selon CNBC, Nvidia et OpenAI négocient un dispositif de garantie financière d’un montant pouvant atteindre 250 milliards de dollars, destiné à sécuriser la construction d’un data center géant de…
Lire l’article complet
IA
ZD Tech : Ce nouvel outil basé sur l'IA pourrait simplifier la vie de vos équipes de conformité (et de vos développeurs)
La scale-up Sumsub frappe un grand coup en ouvrant sa plateforme aux agents d'IA comme ChatGPT et Claude. Une innovation qui transforme de lourdes règles réglementaires en workflows opérationnels. Et aujourd'hui, on décrypte une petite révolution à la frontière de l'intelligence artificielle et de la conformité réglementaire. Imaginez pouvoir traduire une politique complexe de lutte contre le blanchiment d'argent en un workflow opérationnel en seulement quelques minutes. C'est le tour de force réalisé par…
Lire l’article complet
IA
Web pratique : cinq façons d'obtenir davantage de mentions grâce à la recherche basée sur l'IA
Le référencement naturel n'est pas mort ! Voici comment vous pouvez faire remarquer et rendre plus visible votre site Web par ChatGPT, Gemini et autres outils d’intelligence artificielle. Vous perdez discrètement du trafic issu du référencement naturel, même si vos classements n’ont pas changé. Cela signifie-t-il que le référencement naturel traditionnel ne fonctionne plus ? Non. Mais devez-vous pour autant considérer l’IA comme une force terrible qu’il faut simplement accepter et ignorer ? Non…
Lire l’article complet
IA
L'Arcep veut tirer le bilan environnemental des chatbots IA
L'Arcep souhaite collecter les données des entreprises proposant des agents conversationnels afin de mieux évaluer leur impact sur l'environnement. Elle ouvre une consultation publique sur le sujet. Depuis 2020, l'Arcep publie chaque année les résultats de son enquête "Pour un numérique soutenable". Ce rapport publié chaque année offre un tour d'horizon de l'empreinte environnementale du secteur du numérique en France. Et pour ses prochaines éditions, l'Arcep souhaiterait faire évoluer sa formule en collectant les…
Lire l’article complet
IA
LM Studio Bionic adds support for Moonshot AI’s Kimi K3 model
LM Studio’s agentic workflow platform now lets users run tasks with Kimi K3, Moonshot AI’s 2.8-trillion-parameter model with a 1-million-context window. Here are the details. A few days ago, LM Studio announced Bionic, a new app aimed at users looking to leverage open models for agentic tasks. Bionic is designed to handle coding, research, and more complex work involving creating or editing documents and files. Users can run models locally or, when a task…
Lire l’article complet
IA
Anthropic’s Dario Amodei responds: doesn’t oppose open-weight models, but fears Chinese AI
Anthropic founder and CEO Dario Amodei responded on Monday afternoon to murmuring in the industry that his company somehow supports efforts by the U.S. government to ban open-weight Chinese models, or possibly open-weight models generally. “Anyone who has read my past writing should know that I don’t regard such bans as a useful measure, but let me state it clearly so that there is no doubt: Anthropic has never advocated for a ban on…
Lire l’article complet
IA
Microsoft unveils AI security tools it says outperform competing platforms
Microsoft is introducing new AI tools designed to help customers continuously streamline and automate the process of identifying and reducing their exposure to security risks. The new tools come less than a week after OpenAI lost control of two of its security models when they infiltrated the servers of startup Hugging Face. The hack, Hugging Face added, involved “a swarm of tens of thousands of automated actions” that stole internal Hugging Face credentials. The…
Lire l’article complet
IA
Satya Nadella says companies that trust one AI for everything may not survive
On Sunday, Microsoft CEO Satya Nadella doubled down on the shocking warning he issued earlier this month to businesses that use AI, taking it a step further this time. Companies that rely wholly on the proprietary AI labs for their AI needs ultimately won’t survive, he predicts. That’s what he said on CNN’s “Fareed Zakaria GPS.” When Zakaria asked Nadella to explain what constitutes a company sharing too much with an AI model provider,…
Lire l’article complet
IA
Claude, Codex, and other AI tools credited in today’s Apple security releases
Apple’s technical details on the many security fixes included in today’s operating system updates show how quickly AI tools are becoming part of vulnerability research. Here are the details. As soon as the updates were rolled out, Apple updated its security releases page with more information on several of the issues addressed, including the affected components and devices, their potential impact, how Apple fixed them, and the researchers credited with reporting them. One notable…
Lire l’article complet
IA
Daily Brief 'Tuning' could be what Google has planned next for Gemini's morning digest
What you need to know "Tuning" was spotted in a leak conerning Google's Gemini Daily Brief, which could bring specified customization tools to users. A report claims users can tell the AI what they want to see in their morning digests and what they don't. Gemini Spark and Daily Brief were a part of Google's I/O highlights, which turn the AI into a proactive assistant. Google rolled out "daily brief' for Gemini earlier this…
Lire l’article complet
IA
Après Facebook, Instagram et WhatsApp, Meta AI débarque dans les messages privés de Threads
Threads déploie Meta AI dans les messages privés. Les utilisateurs n’ont plus l’obligation de solliciter l’assistant en public. Voici comment ça marche. Les utilisateurs avaient déjà accès à Meta AI dans les messages privés sur Facebook, Instagram et WhatsApp, voilà qu’ils peuvent en avoir l’usage sur Threads, l’alternative à Twitter de l’entreprise. Jusqu’ici, l’intégration de Meta AI dans Threads était partielle et se résumait aux fils de discussions publics. On ne pouvait donc pas…
Lire l’article complet
IA
Prep for Siri AI on Apple Vision Pro with visionOS 26.6
The end-of-cycle updates are often devoid of anything user-facing, but that doesn't mean you should delay updating to visionOS 26.6. Security patches and preparation for Siri AI are worth the update. Not every operating system update is going to have some flashy headlining feature. As we await the OS 27 releases in the fall, the OS 26 cycle is still getting needed updates. Apple has released visionOS 26.6 today alongside the other operating systems…
Lire l’article complet
IA
OpenAI called the Hugging Face attack unprecedented. But we’ve been here before.
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Reading OpenAI’s account last week of how some of its models broke their containment and hacked into the computer systems of Hugging Face, another AI company, was the first time I got genuine chills about what large language models are now able to do. But this is a case of…
Lire l’article complet
IA
Why AI-Driven Cognitive Systems Are Redefining Radar and Electronic Warfare
An overview of how mode-agile threats challenge static library radar/EW systems, and how AI/ML cognitive architectures enable adaptive, real-time countermeasures. What Attendees will Learn Why mode-agile threats render static library systems ineffective — Explore how wartime reserve modes and mode-agile emitters deploy unexpected frequencies, modulation techniques, and hopping schemes that cannot be matched against traditional threat databases, leaving legacy electronic protect, attack, and support systems unable to respond. How AI/ML techniques power cognitive radar/EW…
Lire l’article complet
IA
Why China is giving away its best AI models
Chinese labs like Moonshot are forcing OpenAI, Google, and Anthropic to rethink what they lock away. Chinese labs like Moonshot are forcing OpenAI, Google, and Anthropic to rethink what they lock away. Silicon Valley has spent much of the past week on red alert, digesting the arrival of Moonshot AI’s Kimi K3, a Chinese AI model that can allegedly beat some of the best systems built by US companies at a fraction of the…
Lire l’article complet
IA
Rethinking security for the age of AI
Why security needs a new Cyber Stack — Introducing Project Perception The physics of cybersecurity are changing. Autonomous systems can now reason, adapt and operate continuously. At the same time, the cost of offense is falling, while the volume, velocity and complexity of what must be secured continues to grow. Attackers can generate exploits faster, scale campaigns further and operate with unprecedented efficiency. The approaches built for a world of human actors cannot keep…
Lire l’article complet